Efficient pilot design based on the divide and conquer approach for pilot contamination mitigation in massive MIMO*

Pilot contamination is considered as a bottleneck in the massive multi-input multi-output (MIMO) systems. In a previous work, we proposed a novel iterative algorithm, dubbed channel Zooming Algorithm (CZA), to ensure the mitigation of pilot contamination effect. For this, we assumed that each cell is assigned a set of orthogonal pilot sequences and that these same pilot sequences are reused in other cells. In this paper, we propose a novel strategy of pilot design, based on the divide-and-conquer approach (DCA), aiming to halve the contamination power observed by the channel estimate and to conquer the remaining pilot contamination through the iterations of the CZA. More precisely, instead of being contaminated by a given number of contaminants, as in the initial approach of the CZA, the user, in the proposed approach, is contaminated by a twice the number of contaminants, while halving the pilot contamination power coming from each one of them. We confirm the effectiveness of the proposed scheme with analytical and simulation results and show that the CZA algorithm is more efficient and faster to eliminate the contamination in this case.